1. Field of the Invention
The present invention relates to signal processing, and more particularly, to a method and apparatus of adaptive transmission signal detection based on a signal-to-noise ratio and a Chi-squared distribution for decoding a reception signal of a multi-stream into a transmission signal.
2. Related Art
When a signal is transmitted and received between a transmitter and a receiver in a communication system, the signal may be transmitted or received in the form of a plurality of signal streams. That is, the transmitter such as a base station, a terminal, a relay, etc. generates and transmits the plurality of signal streams, and the plurality of signal streams are received at the receiver through a channel. If the receiver is configured by an apparatus of one or more receiver unit, the plurality of transmitted signal streams are output in the form of a reception signal vector at the receiver for every unit time and input to a detector or a decoder of the receiver.
As a method for detecting an original transmission signal from a plurality of received signals, there have been proposed various methods in consideration of detection error and calculation amount. To obtain the lowest detection error, a maximum likelihood detection may be used. As an algorithm for acquiring the maximum likelihood performance, a sphere decoding may be employed. The sphere decoding may be referred to M. O. Damen and H. El Gamal, “On Maximum-Likelihood Detection and the search for the closest lattice point”, IEEE Trans. Inform. Theory, vol. 49, no. 10, pp. 2389-2402, OCT. 2003 and E. Agrell, T. Errikson, A. Vardy, and K. Zeger, “Closest point search in lattices,” IEEE Trans. Inform. Theory, vol. 48, pp. 2201-2214, August 2002. The sphere decoding guarantees the maximum likelihood performance, but has a shortcoming that the calculation amount is largely varied depending on a condition of channel matrix, particularly a condition number, multiplied to a transmission signal vector or depending on a signal-to-noise ratio (SNR). For example, if the condition number of the channel matrix is approximate to 1 and the SNR is not low, the calculation amount is small as compared with those of other detection methods. On the other hand, if a channel condition is bad or the SNR is low, the calculation amount is so large that a detection time is delayed or an algorithm is not sufficiently performed if there is assigned a limit delay time. The calculation amount in the sphere decoding is generally determined on the basis of an initial radius.
Accordingly, there is a need of a signal processing method in which an algorithm in a sphere decoding and an algorithm of adaptively controlling the calculation amount in the sphere decoding are combined.